Upload files to a Hugging Face repository
Create a large, deduplicated dataset for LLM pre-training
Create datasets with FAQs and SFT prompts
Convert and PR models to Safetensors
Download datasets from a URL
Review and rate queries
Create Reddit dataset
Support by Parquet, CSV, Jsonl, XLS
Launch and explore labeled datasets
Validate JSONL format for fine-tuning
Generate a Parquet file for dataset validation
Convert PDFs to a dataset and upload to Hugging Face
Upload To Hub is a tool designed to simplify the process of uploading files to a Hugging Face repository. It allows users to easily manage and share their datasets, models, and other files directly within the Hugging Face ecosystem. This app is particularly useful for researchers, data scientists, and developers who need to collaborate and distribute their work efficiently.
• File Upload: Supports uploading various file types, including datasets, models, and documentation.
• Multiple File Upload: Enables users to upload multiple files at once, saving time and effort.
• Folder Organization: Allows organization of files into folders for better structure and management.
• Access Control: Provides options to set files as public or private, ensuring data privacy and security.
• Integration with Hugging Face Ecosystem: Seamlessly integrates with Hugging Face tools and libraries.
• Version Control: Includes features for tracking changes and updates to uploaded files.
What types of files can I upload to the Hugging Face Hub?
You can upload a variety of file types, including datasets, models, images, text files, and more. The platform supports popular formats like CSV, JSON, ZIP, and PNG.
How do I set my files to be private?
During the upload process, you can configure access settings. Select the "Private" option to ensure only you or authorized users can access the files.
What if my files are too large to upload?
The Hugging Face Hub has file size limits. If your files exceed these limits, contact support for assistance. You may need to split your files or use alternative upload methods.